Still No Free Lunches: The Price to Pay for Tighter PAC-Bayes Bounds
“No free lunch” results state the impossibility of obtaining meaningful bounds on the error of a learning algorithm without prior assumptions and modelling, which is more or less realistic for a given problem. Some models are “expensive” (strong assumptions, such as sub-Gaussian tails), others are “...
Guardado en:
Autores principales: | , |
---|---|
Formato: | article |
Lenguaje: | EN |
Publicado: |
MDPI AG
2021
|
Materias: | |
Acceso en línea: | https://doaj.org/article/60e63bead773480ead8e98baa4144956 |
Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
id |
oai:doaj.org-article:60e63bead773480ead8e98baa4144956 |
---|---|
record_format |
dspace |
spelling |
oai:doaj.org-article:60e63bead773480ead8e98baa41449562021-11-25T17:30:35ZStill No Free Lunches: The Price to Pay for Tighter PAC-Bayes Bounds10.3390/e231115291099-4300https://doaj.org/article/60e63bead773480ead8e98baa41449562021-11-01T00:00:00Zhttps://www.mdpi.com/1099-4300/23/11/1529https://doaj.org/toc/1099-4300“No free lunch” results state the impossibility of obtaining meaningful bounds on the error of a learning algorithm without prior assumptions and modelling, which is more or less realistic for a given problem. Some models are “expensive” (strong assumptions, such as sub-Gaussian tails), others are “cheap” (simply finite variance). As it is well known, the more you pay, the more you get: in other words, the most expensive models yield the more interesting bounds. Recent advances in robust statistics have investigated procedures to obtain tight bounds while keeping the cost of assumptions minimal. The present paper explores and exhibits what the limits are for obtaining tight probably approximately correct (PAC)-Bayes bounds in a robust setting for cheap models.Benjamin GuedjLouis PujolMDPI AGarticlestatistical learning theoryPAC-Bayes theoryno free lunch theoremsScienceQAstrophysicsQB460-466PhysicsQC1-999ENEntropy, Vol 23, Iss 1529, p 1529 (2021) |
institution |
DOAJ |
collection |
DOAJ |
language |
EN |
topic |
statistical learning theory PAC-Bayes theory no free lunch theorems Science Q Astrophysics QB460-466 Physics QC1-999 |
spellingShingle |
statistical learning theory PAC-Bayes theory no free lunch theorems Science Q Astrophysics QB460-466 Physics QC1-999 Benjamin Guedj Louis Pujol Still No Free Lunches: The Price to Pay for Tighter PAC-Bayes Bounds |
description |
“No free lunch” results state the impossibility of obtaining meaningful bounds on the error of a learning algorithm without prior assumptions and modelling, which is more or less realistic for a given problem. Some models are “expensive” (strong assumptions, such as sub-Gaussian tails), others are “cheap” (simply finite variance). As it is well known, the more you pay, the more you get: in other words, the most expensive models yield the more interesting bounds. Recent advances in robust statistics have investigated procedures to obtain tight bounds while keeping the cost of assumptions minimal. The present paper explores and exhibits what the limits are for obtaining tight probably approximately correct (PAC)-Bayes bounds in a robust setting for cheap models. |
format |
article |
author |
Benjamin Guedj Louis Pujol |
author_facet |
Benjamin Guedj Louis Pujol |
author_sort |
Benjamin Guedj |
title |
Still No Free Lunches: The Price to Pay for Tighter PAC-Bayes Bounds |
title_short |
Still No Free Lunches: The Price to Pay for Tighter PAC-Bayes Bounds |
title_full |
Still No Free Lunches: The Price to Pay for Tighter PAC-Bayes Bounds |
title_fullStr |
Still No Free Lunches: The Price to Pay for Tighter PAC-Bayes Bounds |
title_full_unstemmed |
Still No Free Lunches: The Price to Pay for Tighter PAC-Bayes Bounds |
title_sort |
still no free lunches: the price to pay for tighter pac-bayes bounds |
publisher |
MDPI AG |
publishDate |
2021 |
url |
https://doaj.org/article/60e63bead773480ead8e98baa4144956 |
work_keys_str_mv |
AT benjaminguedj stillnofreelunchesthepricetopayfortighterpacbayesbounds AT louispujol stillnofreelunchesthepricetopayfortighterpacbayesbounds |
_version_ |
1718412287955435520 |